Skill Detail

Deepgram Nova Transcriber

Transcribes audio using the Deepgram Nova-2 API with diarization, punctuation, and smart formatting. Supports streaming via WebSocket and batch via REST with pre-recorded endpoint and callback URLs.

Media & TranscriptionOpenClaw
Media & Transcription OpenClaw Security Reviewed
INSTALL WITH ANY AGENT
npx skills add agentskillexchange/skills --skill deepgram-nova-transcriber-agent Copy
Works best when you want a reusable capability, not another fragile one-off prompt.
At a glance
Author
Deepgram
Last updated
Mar 24, 2026
Quick brief

Deepgram Nova Transcriber integrates with Deepgram’s Nova-2 speech recognition model for high-accuracy transcription. It uses the /v1/listen REST endpoint for batch processing and WebSocket connections at wss://api.deepgram.com/v1/listen for real-time streaming transcription.

How it works

What this skill actually does

The agent configures transcription parameters including diarize=true for speaker identification, punctuate=true for automatic punctuation, smart_format=true for intelligent formatting of dates, numbers, and currency. Supports language detection with detect_language=true across 30+ languages.

For batch processing, it submits audio via URL reference or direct upload with Content-Type: audio/*, using callback URLs for async completion notification. Handles audio formats including WAV, MP3, FLAC, OGG, and WebM. Outputs structured JSON with word-level timestamps, confidence scores, paragraphs, and utterances. Includes post-processing for SRT/VTT subtitle generation and topic detection via Deepgram’s topics=true feature.